OBJECTIVES: To evaluate the efficacy and safety profile of gaboxadol, a selective extrasynaptic GABA(A) agonist (SEGA) previously in development for the treatment of insomnia. METHODS: This was a randomised, double-blind, placebo-controlled, parallel-group, 2-week, Phase III study of gaboxadol 5, 10 and 15mg in outpatients meeting the DSM-IV criteria of primary insomnia (N=742). Zolpidem 10mg was used as active reference. RESULTS: At weeks 1 and 2, significant improvement in total sleep time (sTST) compared to placebo was seen for all doses of gaboxadol (all p<0.05). In addition, gaboxadol 10 and 15mg decreased the number of awakenings (sNAW) (p<0.05) while only gaboxadol 15mg improved wakefulness after sleep onset (sWASO) (p<0.05). At week 1, all doses of gaboxadol significantly improved time-to-sleep onset (sTSO) (p<0.05). At week 2, a sustained effect on sTSO was observed for gaboxadol 15mg. Zolpidem also showed effect on all of these variables. Gaboxadol and zolpidem improved sleep quality, freshness after sleep, daytime function and energy at both weeks. Transient rebound insomnia was observed following discontinuation of treatment with zolpidem, but not gaboxadol. CONCLUSIONS: Gaboxadol 15mg treatment for 2 weeks significantly improved sleep onset and maintenance variables as well as sleep quality and daytime function, as did zolpidem. Gaboxadol 5 and 10mg also showed benefits on most efficacy variables. Gaboxadol was generally safe and well tolerated, with no evidence of withdrawal symptoms or rebound insomnia after discontinuation of short-term treatment. For zolpidem, transient rebound insomnia was observed.